CN117952688A - Classification method and device for merchants and electronic equipment - Google Patents

Abstract

The application discloses a classification method, a classification device and electronic equipment of merchants, wherein the method is applied to the field of big data, the field of financial science and technology or other technical fields, and comprises the following steps: determining index data of different dimensionalities of a merchant to obtain a first index set; performing data preprocessing operation on the first index set to obtain a second index set; calculating the coefficient of each second index in the second index set by adopting a self-adaptive lasso algorithm; normalizing the coefficients of each second index to obtain the weight of each second index; and calculating the evaluation score of the target merchant according to the second index set and the weight of each second index, and classifying the target merchant according to the evaluation score of the target merchant to obtain a classification result. The method and the device solve the problem that in the related art, the merchant is classified by a single angle of transaction payment information, so that the classification result is inaccurate.

Description

Classification method and device for merchants and electronic equipment

Technical Field

The application relates to the field of big data, financial science and technology or other technical fields, in particular to a method and device for classifying merchants and electronic equipment.

Background

With the gradual saturation of the merchant order receiving service, the experience of the merchant customer group is promoted to become an important point of exertion for the financial institution, and the financial institution needs to provide various services and effective management measures for the merchant due to the insufficient market expansion function of the traditional merchant service, so that the service quality of the merchant is improved, the market of the merchant service is expanded, and the experience of the merchant customer group is promoted.

At present, classification management is a common method for managing merchants by a financial institution, but most of the management of the merchants by the financial institution directly applies an enterprise/personal client management method, or classification is carried out on the merchant guest group only from a merchant order-receiving business layer, so that targeted classification management focusing on the main characteristic of the merchants is less realized. Since the merchant service of the financial institution has been upgraded from a single payment service to a comprehensive service covering various services, if the financial institution still evaluates the merchant service and sorts the merchant at a single angle for payment of the settlement commission, an inaccurate sorting result may be caused, so that the financial institution cannot provide the merchant with a suitable service.

Aiming at the problem that in the related art, merchants are classified through a single angle of transaction payment information, so that the classification result is inaccurate, no effective solution is proposed at present.

Disclosure of Invention

The application mainly aims to provide a classification method, a classification device and electronic equipment for merchants, which are used for solving the problem that in the related art, the classification result is inaccurate due to the fact that the merchants are classified through a single angle of transaction payment information.

In order to achieve the above object, according to one aspect of the present application, there is provided a classification method of merchants, the method comprising: determining index data of different dimensionalities of a merchant to obtain a first index set; performing data preprocessing operation on the first index set to obtain a second index set; calculating the coefficient of each second index in the second index set by adopting an adaptive lasso algorithm, wherein the coefficient represents the correlation degree of the second index and the merchant asset; normalizing the coefficients of each second index to obtain the weight of each second index; and calculating the evaluation score of the target merchant according to the second index set and the weight of each second index, and classifying the target merchant according to the evaluation score of the target merchant to obtain a classification result.

Further, determining the index data of different dimensions of the merchant, the obtaining the first index set includes: determining a first category of index set based on transaction information of the merchant, wherein the transaction information at least comprises one of the following information: revenue information, loan information, product sales information; determining a second set of metrics based on information of a transaction object transacting with the merchant, wherein the information of the transaction object includes at least one of: the quantity information of the transaction objects, the transaction amount information of the transaction objects and the recommendation information of the transaction objects; determining a third set of class indicators based on the asset growth information of the merchant; and combining the first type index set, the second type index set and the third type index set to obtain the first index set.

Further, performing a data preprocessing operation on the first index set to obtain a second index set includes: performing outlier processing on the index data in the first index set to obtain a fourth index set; performing missing value processing on the index data in the fourth index set to obtain a fifth index set; and carrying out standardization processing on index data in the fifth index set to obtain the second index set.

Further, performing outlier processing on the index data in the first index set to obtain a fourth index set includes: detecting abnormal index data in the first index set by adopting a preset algorithm, wherein the preset algorithm at least comprises one of the following algorithms: the percentile distribution algorithm and the box graph algorithm, and the abnormal index data refer to index data with values exceeding a preset interval; determining the value range of the abnormal index data according to the service scene to obtain the preset interval; and adjusting the value of the abnormal index data according to the preset interval to obtain the fourth index set.

Further, performing missing value processing on the index data in the fourth index set to obtain a fifth index set includes: acquiring index missing data in the fourth index set, and calculating the missing rate of the index missing data, wherein the index missing data refers to index data with an empty value; deleting the index missing data under the condition that the missing rate exceeds a preset missing rate threshold value; and under the condition that the deletion rate does not exceed the preset deletion rate threshold value, interpolating the index deletion data by adopting a multiple interpolation method to obtain the fifth index set.

Further, calculating coefficients of each second index in the second set of indices using an adaptive lasso algorithm includes: constructing an estimation model of the asset scale of the merchant by adopting the self-adaptive lasso algorithm to obtain a target model, wherein the target model comprises index data in the second index set; constructing constraint conditions of the target model according to the self-adaptive lasso algorithm to obtain target constraint conditions; programming the target model and the target constraint condition by adopting a programming function to obtain codes of the target model; and solving the target model through codes of the target model to obtain coefficients of each second index.

Further, calculating an evaluation score of the target merchant according to the second index set and the weight of each second index, and classifying the target merchant according to the evaluation score of the target merchant, wherein the step of obtaining a classification result comprises the following steps: quantifying the index data in the second index set according to a preset quantification rule to obtain an evaluation score of each second index; calculating the evaluation score of the target merchant according to the evaluation score of each second index and the weight of each second index; and determining a classification result corresponding to the evaluation score of the target merchant according to a preset classification rule to obtain the classification result.

Further, after classifying the target merchant according to the evaluation score of the target merchant to obtain a classification result, the method further includes: determining a recommendation strategy corresponding to the classification result according to a preset recommendation strategy to obtain a target recommendation strategy, wherein the preset recommendation strategy at least comprises one of the following strategies: a recommended strategy and a maintenance strategy, wherein the recommended strategy at least comprises one of the following information: system information, cost compensation information, transaction amount information, recommended activity information and financial service information, wherein the maintenance strategy at least comprises one of the following information: visit information and lecture information; and processing the target merchant according to the target recommendation strategy.

In order to achieve the above object, according to another aspect of the present application, there is provided a classification device for merchants, the device comprising: the first determining unit is used for determining index data of different dimensions of the merchant to obtain a first index set; the first processing unit is used for carrying out data preprocessing operation on the first index set to obtain a second index set; a first computing unit configured to calculate a coefficient of each second index in the second index set using an adaptive lasso algorithm, where the coefficient characterizes a degree of correlation of the second index with a merchant asset; the second processing unit is used for carrying out normalization processing on the coefficient of each second index to obtain the weight of each second index; and the second calculation unit is used for calculating the evaluation score of the target merchant according to the second index set and the weight of each second index, and classifying the target merchant according to the evaluation score of the target merchant to obtain a classification result.

Further, the determining unit includes: a first determining subunit, configured to determine a first type of index set based on transaction information of the merchant, where the transaction information includes at least one of the following information: revenue information, loan information, product sales information; a second determining subunit, configured to determine a second category indicator set based on information of a transaction object that performs a transaction with the merchant, where the information of the transaction object includes at least one of the following information: the quantity information of the transaction objects, the transaction amount information of the transaction objects and the recommendation information of the transaction objects; a third determining subunit configured to determine a third set of class indicators based on the asset growth information of the merchant; and the combining subunit is used for combining the first type index set, the second type index set and the third type index set to obtain the first index set.

Further, the first processing unit includes: the first processing subunit is used for carrying out abnormal value processing on the index data in the first index set to obtain a fourth index set; the second processing subunit is used for carrying out missing value processing on the index data in the fourth index set to obtain a fifth index set; and the third processing subunit is used for carrying out standardization processing on the index data in the fifth index set to obtain the second index set.

Further, the first processing subunit includes: the detection module is used for detecting abnormal index data in the first index set by adopting a preset algorithm, wherein the preset algorithm at least comprises one of the following algorithms: the percentile distribution algorithm and the box graph algorithm, and the abnormal index data refer to index data with values exceeding a preset interval; the determining module is used for determining the value range of the abnormal index data according to the service scene to obtain the preset interval; and the adjusting module is used for adjusting the value of the abnormal index data according to the preset interval to obtain the fourth index set.

Further, the second processing subunit includes: the acquisition module is used for acquiring index missing data in the fourth index set and calculating the missing rate of the index missing data, wherein the index missing data is index data with an empty value; the deletion module is used for deleting the index missing data under the condition that the missing rate exceeds a preset missing rate threshold value; and the interpolation module is used for interpolating the index missing data by adopting a multiple interpolation method under the condition that the missing rate does not exceed the preset missing rate threshold value to obtain the fifth index set.

Further, the first computing unit includes: the first construction subunit is used for constructing an estimation model of the asset scale of the merchant by adopting the self-adaptive lasso algorithm to obtain a target model, wherein the target model comprises index data in the second index set; the second construction subunit is used for constructing constraint conditions of the target model according to the self-adaptive lasso algorithm to obtain target constraint conditions; a programming subunit, configured to program the target model and the target constraint condition by using a programming function, so as to obtain a code of the target model; and the solving subunit is used for solving the target model through codes of the target model to obtain coefficients of each second index.

Further, the second calculation unit includes: the quantization subunit is used for quantizing the index data in the second index set according to a preset quantization rule to obtain an evaluation score of each second index; the calculating subunit is used for calculating the evaluation score of the target merchant according to the evaluation score of each second index and the weight of each second index; and the fourth determining subunit is used for determining a classification result corresponding to the evaluation score of the target merchant according to a preset classification rule to obtain the classification result.

Further, the apparatus further comprises: the second determining unit is configured to determine, after classifying the target merchant according to the evaluation score of the target merchant to obtain a classification result, a recommendation policy corresponding to the classification result according to a preset recommendation policy, so as to obtain a target recommendation policy, where the preset recommendation policy at least includes one of the following policies: a recommended strategy and a maintenance strategy, wherein the recommended strategy at least comprises one of the following information: system information, cost compensation information, transaction amount information, recommended activity information and financial service information, wherein the maintenance strategy at least comprises one of the following information: visit information and lecture information; and the third processing unit is used for processing the target merchant according to the target recommendation strategy.

In order to achieve the above object, according to one aspect of the present application, there is provided a computer readable storage medium including a stored computer program, wherein the computer program, when run, controls a device in which the computer readable storage medium is located to perform the classification method of the merchant according to any one of the above.

To achieve the above object, according to one aspect of the present application, there is provided an electronic device including one or more processors and a memory for storing one or more programs, wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method for classifying merchants according to any one of the above.

According to the application, the following steps are adopted: determining index data of different dimensionalities of a merchant to obtain a first index set; performing data preprocessing operation on the first index set to obtain a second index set; calculating the coefficient of each second index in the second index set by adopting an adaptive lasso algorithm, wherein the coefficient represents the correlation degree of the second index and the merchant asset; normalizing the coefficients of each second index to obtain the weight of each second index; and calculating the evaluation score of the target merchant according to the second index set and the weight of each second index, classifying the target merchant according to the evaluation score of the target merchant to obtain a classification result, and solving the problem of inaccurate classification result caused by classifying the merchant through a single angle of transaction payment information in the related technology. The index data of different dimensions of the merchant are introduced, so that the extraction of the characteristic of the merchant customer group from multiple angles is facilitated, the accuracy of the classification result of the merchant is improved, meanwhile, the coefficient of each index data is calculated by adopting a self-adaptive lasso algorithm for solving the problem of high dimension data, index data with higher correlation degree with the total asset scale of the merchant are screened, and corresponding weights are distributed, so that the evaluation score of the merchant can be calculated, the merchant is classified according to the evaluation score, and the accuracy of the classification result is improved.

Drawings

The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the application. In the drawings:

FIG. 1 is a flow chart of a method for classifying merchants according to a first embodiment of the present application;

FIG. 2 is a schematic diagram I of an alternative method for classifying merchants according to an embodiment I of the present application;

FIG. 3 is a second schematic diagram of an alternative method for classifying merchants according to the first embodiment of the present application;

FIG. 4 is a third schematic diagram of an alternative method of classifying merchants according to one embodiment of the present application;

FIG. 5 is a schematic diagram IV of an alternative method of classifying merchants according to a first embodiment of the application;

FIG. 6 is a schematic diagram of a classification device for merchants according to a second embodiment of the present application;

fig. 7 is a schematic diagram of a classification electronic device for a merchant according to a fifth embodiment of the application.

Detailed Description

It should be noted that, without conflict, the embodiments of the present application and features of the embodiments may be combined with each other. The application will be described in detail below with reference to the drawings in connection with embodiments.

It should be noted that the method and apparatus for processing a file, the method and apparatus for determining a storage medium and an electronic device according to the present application may be used in the financial and scientific field to improve accuracy of a classification result in classifying a merchant, and may also be used in any field other than the financial and scientific field.

It should be noted that, the user information (including, but not limited to, user equipment information, user personal information, collected index information, collected merchant information, etc.) and the data (including, but not limited to, data for analysis, stored data, displayed data, collected index data, collected merchant data, etc.) related to the present application are information and data authorized by the user or fully authorized by each party, and the collection, use and processing of the related data need to comply with the related laws and regulations and standards of the related area, and are provided with corresponding operation entries for the user to select authorization or rejection.

In order that those skilled in the art will better understand the present application, a technical solution in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in which it is apparent that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present application without making any inventive effort, shall fall within the scope of the present application.

It should be noted that the terms "first," "second," and the like in the description and the claims of the present application and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate in order to describe the embodiments of the application herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

Example 1

The present application will be described with reference to preferred implementation steps, and fig. 1 is a flowchart of a method for classifying merchants according to an embodiment of the present application, and as shown in fig. 1, the method includes the following steps:

Step S101, determining index data of different dimensions of a merchant to obtain a first index set.

In the first embodiment, when the financial institution provides the service to the merchant, the merchant can be classified according to the asset size of the merchant because the merchant with larger asset size generally has a more stable business condition, so as to provide the merchant with the appropriate service. In order to classify the merchants, index data related to information such as business conditions, transaction conditions, and customer group information of the merchants needs to be collected, so that the merchants are classified according to the index data of the merchants.

Step S102, data preprocessing operation is carried out on the first index set, and a second index set is obtained.

In the first embodiment, in order to ensure the accuracy of the classification result, preprocessing operation is required to be performed on the collected index data, so that abnormal data (such as noise data, missing data, etc.) in the index data are reduced, the influence of the abnormal data on the classification result of the merchant is avoided, and the accuracy of the classification result of the merchant is improved.

And step S103, calculating the coefficient of each second index in the second index set by adopting an adaptive lasso algorithm, wherein the coefficient represents the correlation degree of the second index and the merchant asset.

In the first embodiment, in order to improve the accuracy of the classification result, the degree of correlation between different index data and the total asset size of the merchant, that is, the coefficient of each second index data, needs to be considered, so that the index with a lower degree of correlation is prevented from reducing the accuracy of the classification result. Therefore, if the coefficient evaluation model of each index data is constructed by adopting the self-adaptive lasso algorithm, and the solution is carried out, if the coefficient obtained by the solution is higher, the correlation degree between the corresponding index data and the total asset scale of the commercial tenant is higher, and if the coefficient obtained by the solution is 0, the uncorrelation between the corresponding index data and the total asset scale of the commercial tenant is illustrated.

Step S104, carrying out normalization processing on the coefficients of each second index to obtain the weight of each second index.

In the first embodiment, since the sum of index coefficients obtained by the adaptive lasso algorithm may not be 1, normalization processing is required to be performed on the coefficients of each second index to obtain the weight of each second index, so that the sum of index weights is equal to 1, and the accuracy in weight calculation is ensured. The calculation formula of the normalization process may be as shown in formula one,

Wherein, beta _j represents the coefficient of the j-th index (namely the second index), phi _j (j=1, 2, …, p) is the weight of the j-th index after normalization treatment (namely the weight),P is the index number.

Step S105, calculating the evaluation score of the target merchant according to the second index set and the weight of each second index, and classifying the target merchant according to the evaluation score of the target merchant to obtain a classification result.

In the first embodiment, after determining the weight of each second index, the index data of each second index is quantized to obtain the evaluation score of each second index, and then the evaluation score of each second index and the weight of each second index are calculated by adopting a linear weighting method to obtain the evaluation score of the target merchant, and then the target merchant is classified according to the evaluation score of the target merchant to obtain the classification result.

In summary, according to the classification method of the merchant provided by the embodiment of the application, the first index set is obtained by determining index data of different dimensions of the merchant; performing data preprocessing operation on the first index set to obtain a second index set; calculating the coefficient of each second index in the second index set by adopting a self-adaptive lasso algorithm, wherein the coefficient represents the correlation degree of the second index and the merchant asset; normalizing the coefficients of each second index to obtain the weight of each second index; the evaluation score of the target merchant is calculated according to the second index set and the weight of each second index, and the target merchant is classified according to the evaluation score of the target merchant, so that a classification result is obtained, and the problem that the classification result is inaccurate due to the fact that the merchant is classified through a single angle of transaction payment information in the related technology is solved. The index data of different dimensions of the merchant are introduced, so that the extraction of the characteristic of the merchant customer group from multiple angles is facilitated, the accuracy of the classification result of the merchant is improved, meanwhile, the coefficient of each index data is calculated by adopting a self-adaptive lasso algorithm for solving the problem of high dimension data, index data with higher correlation degree with the total asset scale of the merchant are screened, and corresponding weights are distributed, so that the evaluation score of the merchant can be calculated, the merchant is classified according to the evaluation score, and the accuracy of the classification result is improved.

Optionally, in the method for classifying merchants provided in the first embodiment of the present application, determining index data of different dimensions of the merchants includes: determining a first category of index set based on transaction information of a merchant, wherein the transaction information comprises at least one of the following information: revenue information, loan information, product sales information; determining a second set of metrics based on information of a transaction object transacted with the merchant, wherein the information of the transaction object includes at least one of: the method comprises the steps of (1) number information of transaction objects, transaction amount information of the transaction objects and recommendation information of the transaction objects; determining a third set of metrics based on the asset growth information of the merchant; and combining the first class index set, the second class index set and the third class index set to obtain a first index set.

In the first embodiment, in order to classify the merchants, the index and index data of the merchants in multiple dimensions need to be acquired, so that the asset scale of the merchants can be comprehensively evaluated in the multiple dimensions, and the accuracy of the classification result is further improved.

Specifically, the evaluation index is selected from three dimensions of economic condition (i.e. the transaction information), ecological condition (i.e. the information of the transaction object transacted with the merchant) and potential trend (i.e. the asset growth information of the merchant), wherein the total amount of the asset of the merchant is taken as the explained variable (y), and the influence of the remaining index variables x _i on y is positive.

Economic conditions refer to the value that merchants bring directly that can be reflected in revenue payouts, and the first category of metrics can include: revenue information, such as cash revenue and commission revenue of the merchant; loan information, such as loan revenue information; sales information, such as sales revenue information for various products.

The ecological condition refers to the contact condition of the merchant with enterprise clients and individual clients, and mainly shows the capability of the merchant for increasing the individual clients and the purchasing capability of the active clients. Thus, the second class of metrics may include: the transaction amount of the merchant, the newly added amount of the customer of the merchant, the transaction amount of the merchant, the recommended activity amount of the merchant to the customer, and the participation activity amount of the customer.

The potential trend refers to an index for predicting economic benefit brought by a merchant in the future, and the third type of index can include: asset retention, number of revenue sources, and total asset growth rate.

Optionally, in the method for classifying merchants provided in the first embodiment of the present application, performing a data preprocessing operation on the first index set to obtain the second index set includes: performing outlier processing on the index data in the first index set to obtain a fourth index set; performing missing value processing on index data in the fourth index set to obtain a fifth index set; and carrying out standardization processing on the index data in the fifth index set to obtain a second index set.

In the first embodiment, since the collected index data has the problems of data missing, inconsistent, noise pollution and the like, the abnormal value processing, missing value processing and standardization processing are required to be performed on the index data in the first index set to obtain the second index set, so that the quality of the index data is improved, the calculation accuracy is improved, and the effect of improving the accuracy of the classification result is achieved.

Specifically, firstly, abnormal value processing is performed on index data to reduce noise in the index data, then, missing value processing is performed on the index data to fill in missing values, so that the problem of inaccurate calculation results caused by the missing values is avoided, and finally, data normalization processing is performed after the data processing, and due to the fact that the collected index data have different units and dimensions, for example, some index data are in units of 'meta', and other index data are in units of 'number'. Therefore, in order to solve the problem of incomparability between data variables caused by different data dimensions, the index data is standardized by adopting a formula II, the formula II is shown as follows,

Wherein v _ij represents the normalized index data, the value of A is 1, the value of B is 2,J index data representing i-th merchant,/>Minimum value in j index data representing i-th merchant,/>, andRepresents the maximum value in the j index data of the i-th merchant.

Optionally, in the method for classifying merchants provided in the first embodiment of the present application, performing outlier processing on the index data in the first index set to obtain a fourth index set includes: detecting abnormal index data in the first index set by adopting a preset algorithm, wherein the preset algorithm at least comprises one of the following algorithms: the percentile distribution algorithm and the box graph algorithm, and the abnormal index data refer to index data with values exceeding a preset interval; determining the value range of the abnormal index data according to the service scene to obtain a preset interval; and adjusting the value of the abnormal index data according to the preset interval to obtain a fourth index set.

In the first embodiment, the collected index data needs to be subjected to abnormal value detection and abnormal value processing, so that the influence caused by data noise is reduced, the accuracy of the index data is improved, and the effect of improving the accuracy of the classification result of the commercial tenant is achieved.

Specifically, outliers are first detected for data outlier processing using the percentile distribution and the bin graph. The percentile distribution sorts the data in order of size and divides it into 100 equal parts, e.g., the 25 th percentile indicates that 25% of the data is less than or equal to it. Outliers may deviate from the percentile distribution of normal data, for example, if a value in a dataset far exceeds the 99 percentile of other values, then that value is likely an outlier. A Box plot (Box plot) is a statistical chart for displaying the distribution of data, which shows the median, upper and lower quartiles, maximum and minimum values of the data, which is likely to be an outlier if the data point is outside the upper and lower quartiles of the Box plot.

After the abnormal value is detected, determining a statistical result of an index corresponding to the abnormal value according to a service scene or service requirement, and determining a value range of the index according to the statistical result, so that the abnormal value is modified into a numerical value in the value range.

Optionally, in the method for classifying merchants provided in the first embodiment of the present application, performing missing value processing on the index data in the fourth index set to obtain a fifth index set includes: acquiring index missing data in a fourth index set, and calculating the missing rate of the index missing data, wherein the index missing data refers to index data with null values; deleting index missing data under the condition that the missing rate exceeds a preset missing rate threshold value; and under the condition that the deletion rate does not exceed a preset deletion rate threshold value, interpolating the index deletion data by adopting a multiple interpolation method to obtain a fifth index set.

In the first embodiment, the acquired index data needs to be subjected to missing data detection and missing data processing, so that the influence caused by data missing is reduced, the accuracy of the index data is improved, and the effect of improving the accuracy of the classification result of the merchant is further achieved.

Specifically, the distribution of the missing data in the index data of each index in the fourth index set, that is, the missing rate of the index data, for example, the index of the daily product sales income within 30 days is collected, and the index data of 15 days is collected, and the missing rate is 50%. After determining the deletion rate of the index data, judging whether the deletion rate is greater than a preset deletion rate threshold (for example, 99%), deleting the index if the deletion rate of the index data of a certain index is greater than the preset deletion rate threshold, and interpolating the deletion data by using a multiple interpolation method or directly assigning the value of the deletion data to 0 if the deletion rate of the index data of the certain index is less than or equal to the preset deletion rate threshold.

Optionally, in the method for classifying merchants provided in the first embodiment of the present application, calculating the coefficient of each second index in the second index set by using an adaptive lasso algorithm includes: constructing an estimation model of the asset scale of the merchant by adopting a self-adaptive lasso algorithm to obtain a target model, wherein the target model comprises index data in a second index set; constructing constraint conditions of a target model according to the self-adaptive lasso algorithm to obtain target constraint conditions; programming the target model and the target constraint condition by adopting a programming function to obtain a code of the target model; and solving the target model through codes of the target model to obtain coefficients of each second index.

In the first embodiment, in order to determine, from the collected index data, index data having a higher degree of correlation with the scale of the merchant asset, an adaptive Lasso (Lasso) algorithm may be used to construct an estimation model of the coefficient of the index weight and a constraint condition of the coefficient of the index weight, so as to determine, according to the estimation model and the constraint condition, index data having a higher degree of correlation with the scale of the merchant asset in the second index set, that is, the coefficient of each second index.

Because the multi-dimensional index variable is selected in the scheme, the index variables are related to each other, and multiple collineation problems are often generated during calculation. When multiple collinearity exists, the coefficient estimation is inaccurate, so that the result is unstable, and the economic meaning of the coefficient estimation value is not easy to explain. Therefore, the scheme selects the adaptive Lasso method which is commonly used for processing the high-dimensional data problem to select variables.

Specifically, the adaptive Lasso assigns different coefficients and weights to different indexes, wherein the coefficient represents the degree of correlation between each index and the asset scale of the merchant, and a coefficient of 0 represents that the index is irrelevant, and index data with the coefficient within a preset range is used for calculation when the merchant is evaluated, so that important index variables are reserved, and unimportant index variables are discarded. The estimation model of the coefficients constructed based on the adaptive Lasso algorithm (i.e. the target model described above) may be as shown in equation three,

Wherein omega _j represents the weight vector of the j index, lambda is more than or equal to 0 as the adjusting parameter,The variable alpha and the variable beta are valued when the third formula is minimum, N represents the number of merchants, y _i represents the total asset scale of the ith merchant, alpha represents the variable in the adaptive Lasso algorithm, p represents the number of indexes in the second index set, x _ij represents the jth index data of the ith merchant, and beta _j represents the coefficient of the jth index.

According to the designed coefficient estimation model and the self-adaptive Lasso related theory, a research model of the bank merchant guest group asset scale can be constructed, as shown in a formula IV,

E＝α+β₁A1+β₂A2+β₃A3+β₄A4+β₅A5+β₆A6+β₇B1+β₈B2+

Beta ₉B3+β₁₀B4+β₁₁C1+β₁₂C2+β₁₃ C3 (fourth)

Wherein E represents the total asset size of the merchant, alpha represents a variable in the formula III, beta _j represents a coefficient of a j-th index, A1 to A6 represent indexes in the first type of indexes, B1 to B4 represent indexes in the second type of indexes, and C1 to C3 represent indexes in the third type of indexes. According to the self-adaptive Lasso theory, adding constraint conditionsWherein/>I=1, 2, …, p, γ >0, s equals 1, where/>Representing the estimated value of the j-th index weight,/>Representing the estimated value of the j-th index coefficient, so that the coefficient satisfies the formula five, which is shown below,

Where Y _i (i=1, 2, …, n) is an observation of the total asset size of the ith merchant, and X _i1,X_i2,…,X_i3 (i=1, 2, …, n) is an observation of the jth index data of the ith merchant.

After constraint conditions of the target model and the coefficients are determined, msgps packages (namely programming functions) in R statistical software are adopted for programming, codes of the target model and the constraint conditions are obtained, the compiled codes are operated for solving, a coefficient path diagram of the self-adaptive Lasso is obtained, index variables and coefficient values of the index variables can be screened out finally, and the higher the coefficient value of the index variable is, the higher the correlation degree of the index variable and the total asset information of the commercial tenant is indicated.

Optionally, in the method for classifying merchants provided in the first embodiment of the present application, calculating an evaluation score of a target merchant according to the second index set and the weight of each second index, and classifying the target merchant according to the evaluation score of the target merchant, where obtaining the classification result includes: quantifying the index data in the second index set according to a preset quantification rule to obtain an evaluation score of each second index; calculating the evaluation score of the target merchant according to the evaluation score of each second index and the weight of each second index; and determining a classification result corresponding to the evaluation score of the target merchant according to a preset classification rule to obtain the classification result.

After calculating the index weight to obtain the weight of the second index in the first embodiment, each index data needs to be quantized to obtain the evaluation score of each index data, and then the evaluation score of the merchant is obtained by calculating according to the evaluation score of each index data and the weight of each index data, so that the merchant is classified. By quantifying index data of the merchant, the evaluation score of each index can be obtained, so that the classification of the merchant is facilitated, the problem of inaccurate evaluation results caused by artificial scoring is avoided, and the accuracy of classification results is improved.

Specifically, the value range of the index data corresponding to the evaluation score of each index of the merchant is determined according to the service scene or the service requirement, that is, the preset quantization rule may be shown in table 1, where when the index data corresponding to the index 1 is 2 ten thousand yuan, the evaluation score of the index data is 2 points, and the index 2, the index 3, the index 4 and the index 5 are examples of the value ranges of other indexes.

TABLE 1

Then, according to the evaluation scores of the various indexes of the commercial tenant, the weight values (namely the weights) of the indexes are multiplied respectively, and the products are added to obtain the evaluation score of the commercial tenant, the calculation formula is shown as a formula six,

F=xf1+yf2+zf3 (six)

Wherein F represents the evaluation score of the merchant,Phi _j represents the weight of the j-th index, F1 contains the indices of the first class of indices, F2 contains the indices of the second class of indices, and F3 contains the indices of the third class of indices.

Finally, comparing the evaluation score of the target merchant with a preset classification rule to determine a classification result of the target merchant, for example, according to the upper limit of the evaluation score of the target merchant being 5 and the lowest being 0, the target merchant can be divided into five classes of clients, wherein the first class of clients is the evaluation score in the range of [0,1], the second class of clients is the evaluation score in the range of (1, 2), the third class of clients is the evaluation score in the range of (2, 3), the fourth class of clients is the evaluation score in the range of (3, 4), and the fifth class of clients is the evaluation score in the range of (4, 5).

Optionally, in the method for classifying merchants provided in the first embodiment of the present application, after classifying the target merchants according to the evaluation scores of the target merchants to obtain classification results, the method further includes: determining a recommendation strategy corresponding to the classification result according to a preset recommendation strategy to obtain a target recommendation strategy, wherein the preset recommendation strategy at least comprises one of the following strategies: a recommended strategy and a maintenance strategy, wherein the recommended strategy at least comprises one of the following information: system information, fee compensation information, transaction amount information, recommended activity information, financial service information, and maintenance policy at least includes one of the following information: visit information and lecture information; and processing the target merchant according to the target recommendation strategy.

After the classification result of the target merchant is determined in the first embodiment, the financial institution is facilitated to provide appropriate services for each class of merchants according to the classification result of the target merchant, the capability of classification planning and targeted services is improved, and the effects of individuation, differentiation and applicability of the services are achieved.

Specifically, classification management is performed according to the characteristics of the commercial tenant groups of each class, wherein the classification management comprises differential recommendation and differential maintenance, the differential recommendation is to provide different recommendation strategies for commercial tenants of different classes, and the differential maintenance is to provide different maintenance strategies for commercial tenants of different classes.

Differentiated recommendation strategies include: examples of recommendation policies may be as shown in table 2, whether to provide a management system, commission subsidized credit, whether to include credit notification, whether to include drainage activity, whether to include proprietary financial services, and the like.

TABLE 2

Differentiated maintenance strategies include: frequency of visit maintenance, community management, thematic equity activities (e.g., flower arrangement activities, etc.), financial training (e.g., developing lectures), and the like. An example of a maintenance policy may be as shown in table 3.

TABLE 3 Table 3

After determining the recommended policy and maintenance policy (i.e., the target recommended policy described above) for the different classes of merchants, differentiated services are performed for each class of merchants.

Alternatively, in an alternative embodiment, the process of determining the target recommendation policy of the target merchant according to the present embodiment may be as shown in fig. 2. Selecting merchant related index variables from different dimensions to form an initial sample set; performing data cleaning and data normalization processing on the initial sample set to obtain a target sample; performing variable selection on a target sample by using a self-adaptive Lasso method to obtain an important index and an index weight; updating the calculated index weight to obtain normalized weight values of different dimension indexes so as to construct a comprehensive value model of the commercial tenant group; scoring the value range of each index, multiplying the score of the commercial tenant group on each evaluation index by the normalized weight value of the index by using a linear weighting method, and carrying out quantization treatment to obtain a comprehensive value score; and dividing the merchant customer group hierarchy according to the comprehensive value score, determining different star grades, and carrying out classification planning and targeted recommendation on different merchants.

Alternatively, in an alternative embodiment, the flow of processing the index data according to the present embodiment may be as shown in fig. 3. And selecting related index variables of the commercial tenant from different dimensions to form an initial sample set, and then carrying out data cleaning and data standardization processing on the initial sample set to obtain a target sample.

Alternatively, in an alternative embodiment, the flow of calculating the weights of the index data according to the present embodiment may be as shown in fig. 4. And constructing a coefficient estimation model of the index data based on the self-adaptive Lasso algorithm, constructing constraint conditions, programming and solving the coefficient estimation model and the constraint conditions, and screening out important index data and index weights of the index data.

Alternatively, in an alternative embodiment, the process of classifying merchants and determining the target recommendation policy according to the present embodiment may be as shown in fig. 5. Normalizing the calculated index, and constructing a comprehensive value model of the commercial tenant group according to each index data and the weight of each index data; quantizing the index data of each index according to a preset quantization rule to obtain an evaluation score of each index; multiplying the evaluation score of the merchant group on each index by the normalized weight value of the index by using a linear weighting method to obtain the comprehensive value score of the target merchant (namely the evaluation score of the target merchant); and dividing the customer group hierarchy of the target merchants according to the comprehensive value scores, determining different classification levels, and carrying out classification planning and targeted management on different merchants.

It should be noted that the steps illustrated in the flowcharts of the figures may be performed in a computer system such as a set of computer executable instructions, and that although a logical order is illustrated in the flowcharts, in some cases the steps illustrated or described may be performed in an order other than that illustrated herein.

Example two

The second embodiment of the present application also provides a classification device for merchants, which needs to be described that the classification device for merchants of the second embodiment of the present application may be used to execute the classification method for merchants provided in the first embodiment of the present application. The classifying device for merchants provided in the second embodiment of the present application is described below.

Fig. 6 is a schematic diagram of a classification device for merchants according to a second embodiment of the application. As shown in fig. 6, the apparatus includes: a first determination unit 601, a first processing unit 602, a first calculation unit 603, a second processing unit 604, and a second calculation unit 605.

Specifically, the first determining unit 601 is configured to determine index data of different dimensions of the merchant, so as to obtain a first index set.

The first processing unit 602 is configured to perform a data preprocessing operation on the first index set to obtain a second index set.

The first calculating unit 603 is configured to calculate a coefficient of each second index in the second index set by using an adaptive lasso algorithm, where the coefficient characterizes a degree of correlation between the second index and the merchant asset.

And the second processing unit 604 is configured to normalize the coefficient of each second index to obtain a weight of each second index.

The second calculating unit 605 is configured to calculate an evaluation score of the target merchant according to the second index set and the weight of each second index, and classify the target merchant according to the evaluation score of the target merchant, so as to obtain a classification result.

According to the classifying device for the merchants, provided by the embodiment of the application, index data of different dimensions of the merchants are determined through the first determining unit 601, so that a first index set is obtained; the first processing unit 602 performs data preprocessing operation on the first index set to obtain a second index set; the first computing unit 603 calculates a coefficient of each second index in the second index set by adopting an adaptive lasso algorithm, wherein the coefficient represents a correlation degree between the second index and the merchant asset; the second processing unit 604 performs normalization processing on the coefficient of each second index to obtain the weight of each second index; the second calculating unit 605 calculates the evaluation score of the target merchant according to the second index set and the weight of each second index, classifies the target merchant according to the evaluation score of the target merchant to obtain a classification result, and solves the problem that the classification result is inaccurate due to the fact that the merchant is classified through a single angle of transaction payment information in the related art. The index data of different dimensions of the merchant are introduced, so that the extraction of the characteristic of the merchant customer group from multiple angles is facilitated, the accuracy of the classification result of the merchant is improved, meanwhile, the coefficient of each index data is calculated by adopting a self-adaptive lasso algorithm for solving the problem of high dimension data, index data with higher correlation degree with the total asset scale of the merchant are screened, and corresponding weights are distributed, so that the evaluation score of the merchant can be calculated, the merchant is classified according to the evaluation score, and the accuracy of the classification result is improved.

Optionally, in the classifying device for merchants according to the second embodiment of the present application, the first determining unit 601 includes: a first determining subunit, configured to determine a first type of index set based on transaction information of the merchant, where the transaction information includes at least one of the following information: revenue information, loan information, product sales information; a second determining subunit, configured to determine a second category indicator set based on information of a transaction object that performs a transaction with the merchant, where the information of the transaction object includes at least one of the following information: the method comprises the steps of (1) number information of transaction objects, transaction amount information of the transaction objects and recommendation information of the transaction objects; a third determining subunit configured to determine a third set of class indicators based on the asset growth information of the merchant; and the combination subunit is used for combining the first category index set, the second category index set and the third category index set to obtain the first index set.

Optionally, in the classifying device for merchants according to the second embodiment of the present application, the first processing unit 602 includes: the first processing subunit is used for performing outlier processing on the index data in the first index set to obtain a fourth index set; the second processing subunit is used for carrying out missing value processing on the index data in the fourth index set to obtain a fifth index set; and the third processing subunit is used for carrying out standardization processing on the index data in the fifth index set to obtain a second index set.

Optionally, in the classifying device for merchants provided in the second embodiment of the present application, the first processing subunit includes: the detection module is used for detecting abnormal index data in the first index set by adopting a preset algorithm, wherein the preset algorithm at least comprises one of the following algorithms: the percentile distribution algorithm and the box graph algorithm, and the abnormal index data refer to index data with values exceeding a preset interval; the determining module is used for determining the value range of the abnormal index data according to the service scene to obtain a preset interval; the adjusting module is used for adjusting the value of the abnormal index data according to the preset interval to obtain a fourth index set.

Optionally, in the classifying device for merchants provided in the second embodiment of the present application, the second processing subunit includes: the acquisition module is used for acquiring index missing data in the fourth index set and calculating the missing rate of the index missing data, wherein the index missing data refers to index data with an empty value; the deletion module is used for deleting index deletion data under the condition that the deletion rate exceeds a preset deletion rate threshold value; and the interpolation module is used for interpolating the index missing data by adopting a multiple interpolation method under the condition that the missing rate does not exceed the preset missing rate threshold value to obtain a fifth index set.

Optionally, in the classifying device for merchants according to the second embodiment of the present application, the first computing unit 603 includes: the first construction subunit is used for constructing an estimation model of the asset scale of the commercial tenant by adopting a self-adaptive lasso algorithm to obtain a target model, wherein the target model comprises index data in a second index set; the second construction subunit is used for constructing constraint conditions of the target model according to the self-adaptive lasso algorithm to obtain target constraint conditions; the programming subunit is used for programming the target model and the target constraint condition by adopting a programming function to obtain codes of the target model; and the solving subunit is used for solving the target model through codes of the target model to obtain coefficients of each second index.

Optionally, in the classifying device for merchants according to the second embodiment of the present application, the second calculating unit 605 includes: the quantization subunit is used for quantizing the index data in the second index set according to a preset quantization rule to obtain an evaluation score of each second index; the calculating subunit is used for calculating the evaluation score of the target merchant according to the evaluation score of each second index and the weight of each second index; and the fourth determining subunit is used for determining a classification result corresponding to the evaluation score of the target merchant according to a preset classification rule to obtain the classification result.

Optionally, in the classifying device for merchants provided in the second embodiment of the present application, the device further includes: the second determining unit is configured to determine a recommendation policy corresponding to the classification result according to a preset recommendation policy after classifying the target merchant according to the evaluation score of the target merchant to obtain the classification result, so as to obtain the target recommendation policy, where the preset recommendation policy at least includes one of the following policies: a recommended strategy and a maintenance strategy, wherein the recommended strategy at least comprises one of the following information: system information, fee compensation information, transaction amount information, recommended activity information, financial service information, and maintenance policy at least includes one of the following information: visit information and lecture information; and the third processing unit is used for processing the target merchant according to the target recommendation strategy.

The classifying device of the merchant comprises a processor and a memory, wherein the first determining unit 601, the first processing unit 602, the first calculating unit 603, the second processing unit 604, the second calculating unit 605 and the like are all stored in the memory as program units, and the processor executes the program units stored in the memory to realize corresponding functions.

The processor includes a kernel, and the kernel fetches the corresponding program unit from the memory. The kernel can be provided with one or more than one, and the accuracy of the classification result is improved by adjusting the kernel parameters.

The memory may include volatile memory, random Access Memory (RAM), and/or nonvolatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM), among other forms in computer readable media, the memory including at least one memory chip.

A third embodiment of the present invention provides a computer-readable storage medium having a program stored thereon, which when executed by a processor, implements a method for classifying merchants.

The fourth embodiment of the invention provides a processor, which is used for running a program, wherein the program runs to execute a classification method of merchants.

As shown in fig. 7, a fifth embodiment of the present invention provides an electronic device, where the device includes a processor, a memory, and a program stored in the memory and executable on the processor, and the processor implements the following steps when executing the program: determining index data of different dimensionalities of a merchant to obtain a first index set; performing data preprocessing operation on the first index set to obtain a second index set; calculating the coefficient of each second index in the second index set by adopting a self-adaptive lasso algorithm, wherein the coefficient represents the correlation degree of the second index and the merchant asset; normalizing the coefficients of each second index to obtain the weight of each second index; and calculating the evaluation score of the target merchant according to the second index set and the weight of each second index, and classifying the target merchant according to the evaluation score of the target merchant to obtain a classification result.

The processor also realizes the following steps when executing the program: determining index data of different dimensions of the merchant, the obtaining a first index set includes: determining a first category of index set based on transaction information of a merchant, wherein the transaction information comprises at least one of the following information: revenue information, loan information, product sales information; determining a second set of metrics based on information of a transaction object transacted with the merchant, wherein the information of the transaction object includes at least one of: the method comprises the steps of (1) number information of transaction objects, transaction amount information of the transaction objects and recommendation information of the transaction objects; determining a third set of metrics based on the asset growth information of the merchant; and combining the first class index set, the second class index set and the third class index set to obtain a first index set.

The processor also realizes the following steps when executing the program: performing data preprocessing operation on the first index set to obtain a second index set, wherein the step of obtaining the second index set comprises the following steps: performing outlier processing on the index data in the first index set to obtain a fourth index set; performing missing value processing on index data in the fourth index set to obtain a fifth index set; and carrying out standardization processing on the index data in the fifth index set to obtain a second index set.

The processor also realizes the following steps when executing the program: performing outlier processing on the index data in the first index set to obtain a fourth index set includes: detecting abnormal index data in the first index set by adopting a preset algorithm, wherein the preset algorithm at least comprises one of the following algorithms: the percentile distribution algorithm and the box graph algorithm, and the abnormal index data refer to index data with values exceeding a preset interval; determining the value range of the abnormal index data according to the service scene to obtain a preset interval; and adjusting the value of the abnormal index data according to the preset interval to obtain a fourth index set.

The processor also realizes the following steps when executing the program: performing missing value processing on the index data in the fourth index set to obtain a fifth index set, wherein the obtaining the fifth index set comprises the following steps: acquiring index missing data in a fourth index set, and calculating the missing rate of the index missing data, wherein the index missing data refers to index data with null values; deleting index missing data under the condition that the missing rate exceeds a preset missing rate threshold value; and under the condition that the deletion rate does not exceed a preset deletion rate threshold value, interpolating the index deletion data by adopting a multiple interpolation method to obtain a fifth index set.

The processor also realizes the following steps when executing the program: the calculating coefficients of each second index in the second set of indices using the adaptive lasso algorithm includes: constructing an estimation model of the asset scale of the merchant by adopting a self-adaptive lasso algorithm to obtain a target model, wherein the target model comprises index data in a second index set; constructing constraint conditions of a target model according to the self-adaptive lasso algorithm to obtain target constraint conditions; programming the target model and the target constraint condition by adopting a programming function to obtain a code of the target model; and solving the target model through codes of the target model to obtain coefficients of each second index.

The processor also realizes the following steps when executing the program: calculating the evaluation score of the target merchant according to the second index set and the weight of each second index, classifying the target merchant according to the evaluation score of the target merchant, and obtaining the classification result comprises the following steps: quantifying the index data in the second index set according to a preset quantification rule to obtain an evaluation score of each second index; calculating the evaluation score of the target merchant according to the evaluation score of each second index and the weight of each second index; and determining a classification result corresponding to the evaluation score of the target merchant according to a preset classification rule to obtain the classification result.

The processor also realizes the following steps when executing the program: after classifying the target merchants according to the evaluation scores of the target merchants to obtain classification results, the method further comprises the following steps: determining a recommendation strategy corresponding to the classification result according to a preset recommendation strategy to obtain a target recommendation strategy, wherein the preset recommendation strategy at least comprises one of the following strategies: a recommended strategy and a maintenance strategy, wherein the recommended strategy at least comprises one of the following information: system information, fee compensation information, transaction amount information, recommended activity information, financial service information, and maintenance policy at least includes one of the following information: visit information and lecture information; and processing the target merchant according to the target recommendation strategy.

The device herein may be a server, PC, PAD, cell phone, etc.

The application also provides a computer program product adapted to perform, when executed on a data processing device, a program initialized with the method steps of: determining index data of different dimensionalities of a merchant to obtain a first index set; performing data preprocessing operation on the first index set to obtain a second index set; calculating the coefficient of each second index in the second index set by adopting a self-adaptive lasso algorithm, wherein the coefficient represents the correlation degree of the second index and the merchant asset; normalizing the coefficients of each second index to obtain the weight of each second index; and calculating the evaluation score of the target merchant according to the second index set and the weight of each second index, and classifying the target merchant according to the evaluation score of the target merchant to obtain a classification result.

When executed on a data processing device, is further adapted to carry out a program initialized with the method steps of: determining index data of different dimensions of the merchant, the obtaining a first index set includes: determining a first category of index set based on transaction information of a merchant, wherein the transaction information comprises at least one of the following information: revenue information, loan information, product sales information; determining a second set of metrics based on information of a transaction object transacted with the merchant, wherein the information of the transaction object includes at least one of: the method comprises the steps of (1) number information of transaction objects, transaction amount information of the transaction objects and recommendation information of the transaction objects; determining a third set of metrics based on the asset growth information of the merchant; and combining the first class index set, the second class index set and the third class index set to obtain a first index set.

When executed on a data processing device, is further adapted to carry out a program initialized with the method steps of: performing data preprocessing operation on the first index set to obtain a second index set, wherein the step of obtaining the second index set comprises the following steps: performing outlier processing on the index data in the first index set to obtain a fourth index set; performing missing value processing on index data in the fourth index set to obtain a fifth index set; and carrying out standardization processing on the index data in the fifth index set to obtain a second index set.

When executed on a data processing device, is further adapted to carry out a program initialized with the method steps of: performing outlier processing on the index data in the first index set to obtain a fourth index set includes: detecting abnormal index data in the first index set by adopting a preset algorithm, wherein the preset algorithm at least comprises one of the following algorithms: the percentile distribution algorithm and the box graph algorithm, and the abnormal index data refer to index data with values exceeding a preset interval; determining the value range of the abnormal index data according to the service scene to obtain a preset interval; and adjusting the value of the abnormal index data according to the preset interval to obtain a fourth index set.

When executed on a data processing device, is further adapted to carry out a program initialized with the method steps of: performing missing value processing on the index data in the fourth index set to obtain a fifth index set, wherein the obtaining the fifth index set comprises the following steps: acquiring index missing data in a fourth index set, and calculating the missing rate of the index missing data, wherein the index missing data refers to index data with null values; deleting index missing data under the condition that the missing rate exceeds a preset missing rate threshold value; and under the condition that the deletion rate does not exceed a preset deletion rate threshold value, interpolating the index deletion data by adopting a multiple interpolation method to obtain a fifth index set.

When executed on a data processing device, is further adapted to carry out a program initialized with the method steps of: the calculating coefficients of each second index in the second set of indices using the adaptive lasso algorithm includes: constructing an estimation model of the asset scale of the merchant by adopting a self-adaptive lasso algorithm to obtain a target model, wherein the target model comprises index data in a second index set; constructing constraint conditions of a target model according to the self-adaptive lasso algorithm to obtain target constraint conditions; programming the target model and the target constraint condition by adopting a programming function to obtain a code of the target model; and solving the target model through codes of the target model to obtain coefficients of each second index.

When executed on a data processing device, is further adapted to carry out a program initialized with the method steps of: calculating the evaluation score of the target merchant according to the second index set and the weight of each second index, classifying the target merchant according to the evaluation score of the target merchant, and obtaining the classification result comprises the following steps: quantifying the index data in the second index set according to a preset quantification rule to obtain an evaluation score of each second index; calculating the evaluation score of the target merchant according to the evaluation score of each second index and the weight of each second index; and determining a classification result corresponding to the evaluation score of the target merchant according to a preset classification rule to obtain the classification result.

When executed on a data processing device, is further adapted to carry out a program initialized with the method steps of: after classifying the target merchants according to the evaluation scores of the target merchants to obtain classification results, the method further comprises the following steps: determining a recommendation strategy corresponding to the classification result according to a preset recommendation strategy to obtain a target recommendation strategy, wherein the preset recommendation strategy at least comprises one of the following strategies: a recommended strategy and a maintenance strategy, wherein the recommended strategy at least comprises one of the following information: system information, fee compensation information, transaction amount information, recommended activity information, financial service information, and maintenance policy at least includes one of the following information: visit information and lecture information; and processing the target merchant according to the target recommendation strategy.

It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

In one typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include volatile memory in a computer-readable medium, random Access Memory (RAM) and/or nonvolatile memory, etc., such as Read Only Memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

Computer readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of storage media for a computer include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. Computer-readable media, as defined herein, does not include transitory computer-readable media (transmission media), such as modulated data signals and carrier waves.

It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises an element.

It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The foregoing is merely exemplary of the present application and is not intended to limit the present application. Various modifications and variations of the present application will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement, etc. which come within the spirit and principles of the application are to be included in the scope of the claims of the present application.

Claims (10)

1. A method for classifying merchants, comprising:

determining index data of different dimensionalities of a merchant to obtain a first index set;

Performing data preprocessing operation on the first index set to obtain a second index set;

calculating the coefficient of each second index in the second index set by adopting an adaptive lasso algorithm, wherein the coefficient represents the correlation degree of the second index and the merchant asset;

Normalizing the coefficients of each second index to obtain the weight of each second index;

And calculating the evaluation score of the target merchant according to the second index set and the weight of each second index, and classifying the target merchant according to the evaluation score of the target merchant to obtain a classification result.

2. The method of claim 1, wherein determining the index data for different dimensions of the merchant to obtain the first set of indices comprises:

determining a first category of index set based on transaction information of the merchant, wherein the transaction information at least comprises one of the following information: revenue information, loan information, product sales information;

Determining a second set of metrics based on information of a transaction object transacting with the merchant, wherein the information of the transaction object includes at least one of: the quantity information of the transaction objects, the transaction amount information of the transaction objects and the recommendation information of the transaction objects;

determining a third set of class indicators based on the asset growth information of the merchant;

And combining the first type index set, the second type index set and the third type index set to obtain the first index set.

3. The method of claim 1, wherein performing a data preprocessing operation on the first set of metrics to obtain a second set of metrics comprises:

performing outlier processing on the index data in the first index set to obtain a fourth index set;

performing missing value processing on the index data in the fourth index set to obtain a fifth index set;

and carrying out standardization processing on index data in the fifth index set to obtain the second index set.

4. The method of claim 3, wherein performing outlier processing on the index data in the first set of indices to obtain a fourth set of indices comprises:

detecting abnormal index data in the first index set by adopting a preset algorithm, wherein the preset algorithm at least comprises one of the following algorithms: the percentile distribution algorithm and the box graph algorithm, and the abnormal index data refer to index data with values exceeding a preset interval;

Determining the value range of the abnormal index data according to the service scene to obtain the preset interval;

and adjusting the value of the abnormal index data according to the preset interval to obtain the fourth index set.

5. The method of claim 3, wherein performing missing value processing on the index data in the fourth index set to obtain a fifth index set comprises:

Acquiring index missing data in the fourth index set, and calculating the missing rate of the index missing data, wherein the index missing data refers to index data with an empty value;

Deleting the index missing data under the condition that the missing rate exceeds a preset missing rate threshold value;

And under the condition that the deletion rate does not exceed the preset deletion rate threshold value, interpolating the index deletion data by adopting a multiple interpolation method to obtain the fifth index set.

6. The method of claim 1, wherein calculating coefficients for each second index in the second set of indices using an adaptive lasso algorithm comprises:

Constructing an estimation model of the asset scale of the merchant by adopting the self-adaptive lasso algorithm to obtain a target model, wherein the target model comprises index data in the second index set;

Constructing constraint conditions of the target model according to the self-adaptive lasso algorithm to obtain target constraint conditions;

programming the target model and the target constraint condition by adopting a programming function to obtain codes of the target model;

and solving the target model through codes of the target model to obtain coefficients of each second index.

7. The method of claim 1, wherein calculating an evaluation score of a target merchant according to the second set of indicators and the weight of each second indicator, and classifying the target merchant according to the evaluation score of the target merchant, and obtaining a classification result comprises:

Quantifying the index data in the second index set according to a preset quantification rule to obtain an evaluation score of each second index;

calculating the evaluation score of the target merchant according to the evaluation score of each second index and the weight of each second index;

And determining a classification result corresponding to the evaluation score of the target merchant according to a preset classification rule to obtain the classification result.

8. The method of claim 1, wherein after classifying the target merchant according to the evaluation score of the target merchant, the method further comprises:

Determining a recommendation strategy corresponding to the classification result according to a preset recommendation strategy to obtain a target recommendation strategy, wherein the preset recommendation strategy at least comprises one of the following strategies: a recommended strategy and a maintenance strategy, wherein the recommended strategy at least comprises one of the following information: system information, cost compensation information, transaction amount information, recommended activity information and financial service information, wherein the maintenance strategy at least comprises one of the following information: visit information and lecture information;

And processing the target merchant according to the target recommendation strategy.

9. A merchant classifying device, comprising:

The first determining unit is used for determining index data of different dimensions of the merchant to obtain a first index set;

The first processing unit is used for carrying out data preprocessing operation on the first index set to obtain a second index set;

a first computing unit configured to calculate a coefficient of each second index in the second index set using an adaptive lasso algorithm, where the coefficient characterizes a degree of correlation of the second index with a merchant asset;

The second processing unit is used for carrying out normalization processing on the coefficient of each second index to obtain the weight of each second index;

And the second calculation unit is used for calculating the evaluation score of the target merchant according to the second index set and the weight of each second index, and classifying the target merchant according to the evaluation score of the target merchant to obtain a classification result.

10. An electronic device comprising one or more processors and a memory for storing one or more programs, wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method of classifying merchants of any of claims 1 to 8.